Scipy: fix some future tests

[markotoplak]

Issue

[71](https://github.com/biolab/orange3/actions/runs/34688489002/job/103539549615?pr=7320#step:8:172)
  File "/home/runner/work/orange3/orange3/.tox/beta/lib/python3.14/site-packages/Orange/statistics/util.py", line 13, in <module>
    import scipy.stats.stats
ModuleNotFoundError: No module named 'scipy.stats.stats'

[codecov]

Codecov Report

❌ Patch coverage is 75.00000% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 88.98%. Comparing base (13db694) to head (ab85537).
⚠️ Report is 17 commits behind head on master.

❌ Your patch check has failed because the patch coverage (75.00%) is below the target coverage (95.00%). You can increase the patch coverage or adjust the target coverage.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #7321      +/-   ##
==========================================
- Coverage   88.98%   88.98%   -0.01%     
==========================================
  Files         336      337       +1     
  Lines       74567    74598      +31     
==========================================
+ Hits        66353    66378      +25     
- Misses       8214     8220       +6     

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[janezd]

Should we try running tests with https://docs.scipy.org/doc/scipy/reference/sparse.migration_to_sparray.html#use-tests-to-find-and-spots?

[markotoplak]

Should we try running tests with https://docs.scipy.org/doc/scipy/reference/sparse.migration_to_sparray.html#use-tests-to-find-and-spots?

Probably...

[janezd]

I tried the above tests. The only warnings I got were related to dimension reduction -- sum and mean over axes on sparse matrices produce 2d matrices, while on sparse arrays they will reduce dimension. Warnings came from our code and from sklear. In all cases, were happy with reduced arrays -- we actually make them dense and ravel them, as does sklearn.

Our code seems ready for eventual migration to sparse arrays, when the time comes.